Author's School

Olin Business School


English (en)

Date of Award

Spring 5-15-2023

Degree Type


Degree Name

Doctor of Philosophy (PhD)

Chair and Committee

Barton Hamilton

Committee Members

Stephen Ryan, Timothy McBride, A Mark Fendrick, Brent Hickman,


This dissertation studies how payment incentives in health care change clinician and patient behavior and how different “value-based” payment models may impact outcomes. In Chapter 1, I explore optimal design of healthcare payment models in the presence of market frictions to targeting benefits. I leverage a natural experiment where an employer introduces Direct Primary Care, a form of care delivery where clinicians are paid a capitated monthly fee for high-touch access to a bundle of primary care services. This model modifies moral and behavioral hazard incentives for both patients and clinicians while changing the trade-off between preventive investments and downstream care. I document selection into Direct Primary Care by younger and less costly employees, who have lower primary care spending and shorter job tenures, suggesting the importance of switching costs. Using an instrumental variables approach that leverages plan inertia and a difference-in-difference strategy, I examine the impact on costs of care and demand for preventive care and low-value cardiac imaging. Patient out-of-pocket costs increase for those choosing the Direct Care Plan, and total costs increase for lower spenders but not for more expensive patients. I also find a decrease in potentially low-value imaging and an increase in high-value mammography screening, suggesting that quality improves.

Chapter 2, joint with Ross Klosterman and Namrata Ramakrishna, provides the first data on Direct Primary Care practices nationally and the characteristics related to practice pricing and location decisions. Average adult price charged by a DPC practice is $81.33 per month. Median income and not accepting children as patients is associated with higher prices. Offering more services is not associated with higher prices, and medication dispensing is associated with lower prices. Lower poverty rate, lower percentages of black residents, and higher education status is associated with more physical locations, with broader implications for access to care and organizational care delivery structure in health care.

In Chapter 3, I examine health system responses to changes in payment incentives, studying the Maryland global hospital budget model, which previous evaluations have shown promise in reducing costs. However, the mechanisms by which improved outcomes were achieved have not been studied. Using a dynamic difference-in-difference model and testing the sensitivity of results to the presence of pre-trends, I examine changes in health system investments following the adoption of the global hospital budget model. I find a decrease in volume, a shift from clinical to non-clinical employees in Maryland relative to other states, and an increase in the administrative share of spending, suggesting a shift to lower-skilled workers and a change in care functions. Heart failure readmissions fall and pneumonia mortality increases, while technology adoption is unchanged.

Available for download on Wednesday, July 10, 2024